Sample RankingRecallPrecisionHmean
SRC-B-TextProcessingLab100.00%100.00%100.00%
SRC-B-MachineLearningLab100.00%100.00%100.00%
HoText_v1100.00%100.00%100.00%
Sogou_MM100.00%100.00%100.00%
Alibaba-PAI91.67%91.67%91.67%
Yunos_Robot1.083.33%100.00%90.91%
Baidu IDL v383.33%100.00%90.91%
Sensetime V183.33%100.00%90.91%
FOTS83.33%100.00%90.91%
e2e-mlt + bug fix83.33%100.00%90.91%
e2e-mlt91.67%84.62%88.00%
Beam search CUNI +S (decoding for TextSpotter - no postprocessing)83.33%90.91%86.96%
Megvii-Image++83.33%90.91%86.96%
Baidu VIS83.33%90.91%86.96%
SRC-B-MachineLearningLab83.33%90.91%86.96%
Baidu VIS v283.33%90.91%86.96%
Baidu VIS v2+83.33%83.33%83.33%
TextProposals + DictNet83.33%76.92%80.00%
Tencent Youtu83.33%76.92%80.00%
CPN (multi-scale)83.33%76.92%80.00%
NVTextSpotter100.00%63.16%77.42%
M_FOTS_Orig75.00%75.00%75.00%
gts-swint-cocoonly-s4-3_175.00%56.25%64.29%
Baidu IDL66.67%47.06%55.17%
Baseline OpenCV3.0 + Tesseract41.67%71.43%52.63%
Char-level PTSD on SynthText58.33%41.18%48.28%
Deep2Text-MO16.67%22.22%19.05%
Beam search CUNI (decoding for TextSpotter - spell checked)8.33%100.00%15.38%
Baseline-TextSpotter0.00%0.00%0.00%